|
1
|
|
|
<?php |
|
2
|
|
|
/** @file |
|
3
|
|
|
* The scanner. |
|
4
|
|
|
*/ |
|
5
|
|
|
|
|
6
|
|
|
namespace QueryPath\CSS; |
|
7
|
|
|
|
|
8
|
|
|
/** |
|
9
|
|
|
* Scanner for CSS selector parsing. |
|
10
|
|
|
* |
|
11
|
|
|
* This provides a simple scanner for traversing an input stream. |
|
12
|
|
|
* |
|
13
|
|
|
* @ingroup querypath_css |
|
14
|
|
|
*/ |
|
15
|
|
|
final class Scanner |
|
16
|
|
|
{ |
|
17
|
|
|
|
|
18
|
|
|
public $is; |
|
19
|
|
|
public $value; |
|
20
|
|
|
public $token; |
|
21
|
|
|
|
|
22
|
|
|
public $recurse = false; |
|
23
|
|
|
public $it = 0; |
|
24
|
|
|
|
|
25
|
|
|
/** |
|
26
|
|
|
* Given a new input stream, tokenize the CSS selector string. |
|
27
|
|
|
* |
|
28
|
|
|
* @see InputStream |
|
29
|
|
|
* @param InputStream $in |
|
30
|
|
|
* An input stream to be scanned. |
|
31
|
|
|
*/ |
|
32
|
|
|
public function __construct(InputStream $in) |
|
33
|
|
|
{ |
|
34
|
|
|
$this->is = $in; |
|
35
|
|
|
} |
|
36
|
|
|
|
|
37
|
|
|
/** |
|
38
|
|
|
* Return the position of the reader in the string. |
|
39
|
|
|
*/ |
|
40
|
|
|
public function position(): int |
|
41
|
|
|
{ |
|
42
|
|
|
return $this->is->position; |
|
43
|
|
|
} |
|
44
|
|
|
|
|
45
|
|
|
/** |
|
46
|
|
|
* See the next char without removing it from the stack. |
|
47
|
|
|
* |
|
48
|
|
|
* @return string |
|
49
|
|
|
* Returns the next character on the stack. |
|
50
|
|
|
*/ |
|
51
|
|
|
public function peek(): string |
|
52
|
|
|
{ |
|
53
|
|
|
return $this->is->peek(); |
|
54
|
|
|
} |
|
55
|
|
|
|
|
56
|
|
|
/** |
|
57
|
|
|
* Get the next token in the input stream. |
|
58
|
|
|
* |
|
59
|
|
|
* This sets the current token to the value of the next token in |
|
60
|
|
|
* the stream. |
|
61
|
|
|
* |
|
62
|
|
|
* @return int |
|
63
|
|
|
* Returns an int value corresponding to one of the Token constants, |
|
64
|
|
|
* or FALSE if the end of the string is reached. (Remember to use |
|
65
|
|
|
* strong equality checking on FALSE, since 0 is a valid token id.) |
|
66
|
|
|
* @throws ParseException |
|
67
|
|
|
* @throws \QueryPath\Exception |
|
68
|
|
|
*/ |
|
69
|
|
|
public function nextToken(): int |
|
70
|
|
|
{ |
|
71
|
|
|
$tok = -1; |
|
72
|
|
|
++$this->it; |
|
73
|
|
|
if ($this->is->isEmpty()) { |
|
74
|
|
|
if ($this->recurse) { |
|
75
|
|
|
throw new \QueryPath\Exception('Recursion error detected at iteration ' . $this->it . '.'); |
|
76
|
|
|
} |
|
77
|
|
|
//print "{$this->it}: All done\n"; |
|
78
|
|
|
$this->recurse = true; |
|
79
|
|
|
$this->token = false; |
|
80
|
|
|
|
|
81
|
|
|
return false; |
|
|
|
|
|
|
82
|
|
|
} |
|
83
|
|
|
$ch = $this->is->consume(); |
|
84
|
|
|
//print __FUNCTION__ . " Testing $ch.\n"; |
|
85
|
|
|
if (ctype_space($ch)) { |
|
86
|
|
|
$this->value = ' '; // Collapse all WS to a space. |
|
87
|
|
|
$this->token = $tok = Token::WHITE; |
|
88
|
|
|
|
|
89
|
|
|
//$ch = $this->is->consume(); |
|
90
|
|
|
return $tok; |
|
91
|
|
|
} |
|
92
|
|
|
|
|
93
|
|
|
if ($ch === '-' || $ch === '_' || ctype_alnum($ch)) { |
|
94
|
|
|
// It's a character |
|
95
|
|
|
$this->value = $ch; //strtolower($ch); |
|
96
|
|
|
$this->token = $tok = Token::CHAR; |
|
97
|
|
|
|
|
98
|
|
|
return $tok; |
|
99
|
|
|
} |
|
100
|
|
|
|
|
101
|
|
|
$this->value = $ch; |
|
102
|
|
|
|
|
103
|
|
|
switch ($ch) { |
|
104
|
|
|
case '*': |
|
105
|
|
|
$tok = Token::STAR; |
|
106
|
|
|
break; |
|
107
|
|
|
case chr(ord('>')): |
|
108
|
|
|
$tok = Token::RANGLE; |
|
109
|
|
|
break; |
|
110
|
|
|
case '.': |
|
111
|
|
|
$tok = Token::DOT; |
|
112
|
|
|
break; |
|
113
|
|
|
case '#': |
|
114
|
|
|
$tok = Token::OCTO; |
|
115
|
|
|
break; |
|
116
|
|
|
case '[': |
|
117
|
|
|
$tok = Token::LSQUARE; |
|
118
|
|
|
break; |
|
119
|
|
|
case ']': |
|
120
|
|
|
$tok = Token::RSQUARE; |
|
121
|
|
|
break; |
|
122
|
|
|
case ':': |
|
123
|
|
|
$tok = Token::COLON; |
|
124
|
|
|
break; |
|
125
|
|
|
case '(': |
|
126
|
|
|
$tok = Token::LPAREN; |
|
127
|
|
|
break; |
|
128
|
|
|
case ')': |
|
129
|
|
|
$tok = Token::RPAREN; |
|
130
|
|
|
break; |
|
131
|
|
|
case '+': |
|
132
|
|
|
$tok = Token::PLUS; |
|
133
|
|
|
break; |
|
134
|
|
|
case '~': |
|
135
|
|
|
$tok = Token::TILDE; |
|
136
|
|
|
break; |
|
137
|
|
|
case '=': |
|
138
|
|
|
$tok = Token::EQ; |
|
139
|
|
|
break; |
|
140
|
|
|
case '|': |
|
141
|
|
|
$tok = Token::PIPE; |
|
142
|
|
|
break; |
|
143
|
|
|
case ',': |
|
144
|
|
|
$tok = Token::COMMA; |
|
145
|
|
|
break; |
|
146
|
|
|
case chr(34): |
|
147
|
|
|
$tok = Token::QUOTE; |
|
148
|
|
|
break; |
|
149
|
|
|
case "'": |
|
150
|
|
|
$tok = Token::SQUOTE; |
|
151
|
|
|
break; |
|
152
|
|
|
case '\\': |
|
153
|
|
|
$tok = Token::BSLASH; |
|
154
|
|
|
break; |
|
155
|
|
|
case '^': |
|
156
|
|
|
$tok = Token::CARAT; |
|
157
|
|
|
break; |
|
158
|
|
|
case '$': |
|
159
|
|
|
$tok = Token::DOLLAR; |
|
160
|
|
|
break; |
|
161
|
|
|
case '@': |
|
162
|
|
|
$tok = Token::AT; |
|
163
|
|
|
break; |
|
164
|
|
|
} |
|
165
|
|
|
|
|
166
|
|
|
|
|
167
|
|
|
// Catch all characters that are legal within strings. |
|
168
|
|
|
if ($tok === -1) { |
|
169
|
|
|
// TODO: This should be UTF-8 compatible, but PHP doesn't |
|
170
|
|
|
// have a native UTF-8 string. Should we use external |
|
171
|
|
|
// mbstring library? |
|
172
|
|
|
|
|
173
|
|
|
$ord = ord($ch); |
|
174
|
|
|
// Characters in this pool are legal for use inside of |
|
175
|
|
|
// certain strings. Extended ASCII is used here, though I |
|
176
|
|
|
// Don't know if these are really legal. |
|
177
|
|
|
if (($ord >= 32 && $ord <= 126) || ($ord >= 128 && $ord <= 255)) { |
|
178
|
|
|
$tok = Token::STRING_LEGAL; |
|
179
|
|
|
} else { |
|
180
|
|
|
throw new ParseException('Illegal character found in stream: ' . $ord); |
|
181
|
|
|
} |
|
182
|
|
|
} |
|
183
|
|
|
|
|
184
|
|
|
$this->token = $tok; |
|
185
|
|
|
|
|
186
|
|
|
return $tok; |
|
187
|
|
|
} |
|
188
|
|
|
|
|
189
|
|
|
/** |
|
190
|
|
|
* Get a name string from the input stream. |
|
191
|
|
|
* A name string must be composed of |
|
192
|
|
|
* only characters defined in Token:char: -_a-zA-Z0-9 |
|
193
|
|
|
*/ |
|
194
|
|
|
public function getNameString() |
|
195
|
|
|
{ |
|
196
|
|
|
$buf = ''; |
|
197
|
|
|
while ($this->token === Token::CHAR) { |
|
198
|
|
|
$buf .= $this->value; |
|
199
|
|
|
$this->nextToken(); |
|
200
|
|
|
} |
|
201
|
|
|
|
|
202
|
|
|
return $buf; |
|
203
|
|
|
} |
|
204
|
|
|
|
|
205
|
|
|
/** |
|
206
|
|
|
* This gets a string with any legal 'string' characters. |
|
207
|
|
|
* See CSS Selectors specification, section 11, for the |
|
208
|
|
|
* definition of string. |
|
209
|
|
|
* |
|
210
|
|
|
* This will check for string1, string2, and the case where a |
|
211
|
|
|
* string is unquoted (Oddly absent from the "official" grammar, |
|
212
|
|
|
* though such strings are present as examples in the spec.) |
|
213
|
|
|
* |
|
214
|
|
|
* Note: |
|
215
|
|
|
* Though the grammar supplied by CSS 3 Selectors section 11 does not |
|
216
|
|
|
* address the contents of a pseudo-class value, the spec itself indicates |
|
217
|
|
|
* that a pseudo-class value is a "value between parenthesis" [6.6]. The |
|
218
|
|
|
* examples given use URLs among other things, making them closer to the |
|
219
|
|
|
* definition of 'string' than to 'name'. So we handle them here as strings. |
|
220
|
|
|
*/ |
|
221
|
|
|
public function getQuotedString() |
|
222
|
|
|
{ |
|
223
|
|
|
if ($this->token === Token::QUOTE || $this->token === Token::SQUOTE || $this->token === Token::LPAREN) { |
|
224
|
|
|
$end = ($this->token === Token::LPAREN) ? Token::RPAREN : $this->token; |
|
225
|
|
|
$buf = ''; |
|
226
|
|
|
$escape = false; |
|
227
|
|
|
|
|
228
|
|
|
$this->nextToken(); // Skip the opening quote/paren |
|
229
|
|
|
|
|
230
|
|
|
// The second conjunct is probably not necessary. |
|
231
|
|
|
while ($this->token !== false && $this->token > -1) { |
|
232
|
|
|
//print "Char: $this->value \n"; |
|
233
|
|
|
if ($this->token == Token::BSLASH && !$escape) { |
|
234
|
|
|
// XXX: The backslash (\) is removed here. |
|
235
|
|
|
// Turn on escaping. |
|
236
|
|
|
//$buf .= $this->value; |
|
237
|
|
|
$escape = true; |
|
238
|
|
|
} elseif ($escape) { |
|
239
|
|
|
// Turn off escaping |
|
240
|
|
|
$buf .= $this->value; |
|
241
|
|
|
$escape = false; |
|
242
|
|
|
} elseif ($this->token === $end) { |
|
243
|
|
|
// At end of string; skip token and break. |
|
244
|
|
|
$this->nextToken(); |
|
245
|
|
|
break; |
|
246
|
|
|
} else { |
|
247
|
|
|
// Append char. |
|
248
|
|
|
$buf .= $this->value; |
|
249
|
|
|
} |
|
250
|
|
|
$this->nextToken(); |
|
251
|
|
|
} |
|
252
|
|
|
|
|
253
|
|
|
return $buf; |
|
254
|
|
|
} |
|
255
|
|
|
} |
|
256
|
|
|
|
|
257
|
|
|
// Get the contents inside of a pseudoClass(). |
|
258
|
|
|
public function getPseudoClassString() |
|
259
|
|
|
{ |
|
260
|
|
|
if ($this->token === Token::QUOTE || $this->token === Token::SQUOTE || $this->token === Token::LPAREN) { |
|
261
|
|
|
$end = ($this->token === Token::LPAREN) ? Token::RPAREN : $this->token; |
|
262
|
|
|
$buf = ''; |
|
263
|
|
|
$escape = false; |
|
264
|
|
|
|
|
265
|
|
|
$this->nextToken(); // Skip the opening quote/paren |
|
266
|
|
|
|
|
267
|
|
|
// The second conjunct is probably not necessary. |
|
268
|
|
|
while ($this->token !== false && $this->token > -1) { |
|
269
|
|
|
//print "Char: $this->value \n"; |
|
270
|
|
|
if ($this->token === Token::BSLASH && !$escape) { |
|
271
|
|
|
// XXX: The backslash (\) is removed here. |
|
272
|
|
|
// Turn on escaping. |
|
273
|
|
|
//$buf .= $this->value; |
|
274
|
|
|
$escape = true; |
|
275
|
|
|
} elseif ($escape) { |
|
276
|
|
|
// Turn off escaping |
|
277
|
|
|
$buf .= $this->value; |
|
278
|
|
|
$escape = false; |
|
279
|
|
|
} // Allow nested pseudoclasses. |
|
280
|
|
|
elseif ($this->token === Token::LPAREN) { |
|
281
|
|
|
$buf .= '('; |
|
282
|
|
|
$buf .= $this->getPseudoClassString(); |
|
283
|
|
|
$buf .= ')'; |
|
284
|
|
|
} elseif ($this->token === $end) { |
|
285
|
|
|
// At end of string; skip token and break. |
|
286
|
|
|
$this->nextToken(); |
|
287
|
|
|
break; |
|
288
|
|
|
} else { |
|
289
|
|
|
// Append char. |
|
290
|
|
|
$buf .= $this->value; |
|
291
|
|
|
} |
|
292
|
|
|
$this->nextToken(); |
|
293
|
|
|
} |
|
294
|
|
|
|
|
295
|
|
|
return $buf; |
|
296
|
|
|
} |
|
297
|
|
|
} |
|
298
|
|
|
|
|
299
|
|
|
/** |
|
300
|
|
|
* Get a string from the input stream. |
|
301
|
|
|
* This is a convenience function for getting a string of |
|
302
|
|
|
* characters that are either alphanumber or whitespace. See |
|
303
|
|
|
* the Token::white and Token::char definitions. |
|
304
|
|
|
* |
|
305
|
|
|
* @deprecated This is not used anywhere in QueryPath. |
|
306
|
|
|
*//* |
|
307
|
|
|
public function getStringPlusWhitespace() { |
|
308
|
|
|
$buf = ''; |
|
309
|
|
|
if($this->token === FALSE) {return '';} |
|
310
|
|
|
while ($this->token === Token::char || $this->token == Token::white) { |
|
311
|
|
|
$buf .= $this->value; |
|
312
|
|
|
$this->nextToken(); |
|
313
|
|
|
} |
|
314
|
|
|
return $buf; |
|
315
|
|
|
}*/ |
|
316
|
|
|
|
|
317
|
|
|
} |
|
318
|
|
|
|